Simulation of Long-term Returns with Stochastic Correlations

This paper focuses on a nonlinear stochastic model for financial simulation and forecasting based on assumptions of multivariate stochastic correlation, with an application to the European market. We present in particular the key elements of a structured hierarchical econometric model that can be used to forecast financial and commodity markets relying on statistical and simulation methods. The investment universe includes money-market, fixed-income, inflation-linked bonds as well as equity and commodity indices. For each such investment opportunity a dedicated statistical model has been developed to generate future return paths describing the uncertainty the investment manager is facing over time.